My research is focused around the general theme of developing Machine Learning algorithms with applications to Biomedical Image Analysis and Computer Vision. Lately, I am particularly interested in learning from limited annotations and making Machine Learning models trustworthy to be able to use them clinical settings. See research page for more details.
I joined ETH Zurich in June 2019. Prior to this, I was working as a Senior Research Engineer in ARM Ltd. where I was developing hardware-efficient image processing methods for improving image quality in mobile displays. In 2017, I obtained my Ph.D. degree from Computer Science and Engineering department, Sabanci University, Turkey under the supervision of Prof. Mujdat Cetin. During my Ph.D., I worked on developing Bayesian methods for object segmentation by exploiting nonparametric shape priors.
- Our new preprint titled “Constrained Optimization to Train Neural Networks on Critical and Under-Represented Classes” is out. [Arxiv] [Code]
- Our new preprint titled “Task-agnostic out-of-distribution detection using kernel density estimation” is out. [Arxiv] [Code]
- Our paper titled “Contrastive learning of global and local features for medical image segmentation with limited annotations” is accepted as oral presentation in Neurips 2020. [Paper] [Code]
- I am selected among the top 10% high-scoring reviewers in Neurips 2020.
- I am selected among the outstanding reviewers in MICCAI 2020.
- Our paper titled “Test-Time Adaptable Neural Networks for Robust Medical Image Segmentation” is accepted in MEDIA, Elsevier. [Paper] [Code]
- Our paper titled “Semi-supervised Task-driven Data Augmentation for Medical Image Segmentation” is accepted in MEDIA, Elsevier. [Paper] [Code]